Density And Diversity Of Nontarget Insects Killed By Suburban Electric Insect Traps

Vol. 107, No. 2, March & April, 1996

77

DENSITY AND DIVERSITY OF NONTARGET INSECTS
KILLED BY SUBURBAN ELECTRIC INSECT TRAPS 1 2

Timothy B. Frick, Douglas W. Tallamy^

ABSTRACT: Our survey of insects electroculed during routine use of electric insect traps revealed
only 31 biting flies, a minute proportion (0.22%) of the 13,789 total insects counted. In contrast,
species from 12 orders and more than 104 nontarget insect families, including 1 ,868 predators and
parasites (13.5%) and 6,670 nonbiting aquatic insects (48.4%) were destroyed. The heavy toll on
nontarget insects and the near absence of biting flies in catches suggests that electric insect traps are
worthless for biting fly reduction — and probably are counterproductive — to homeowners and
other consumers.

Electric insect traps (e.g. t Zapper™, Bugwacker™ and Bug Blaster™; here-
after, “zappers”) use ultraviolet light to lure flying insects toward an electrified
metal grid, where they are destroyed by the thousands on warm summer nights.
Homeowners buy traps to rid their surroundings of annoying biting flies, and
continuous snaps, crackles, and pops emanating from an active zapper seem to
confirm their effectiveness. Traps are commonly used near aquatic habitats,
waterfront areas, toll booths, campgrounds, industrial parks, restaurants, swim-
ming pools, and suburban backyards. In suburban yards, traps are often run
throughout the summer months, some only during the evening hours and some
continually.

Although the target insects are primarily mosquitoes (Culicidae) and no-
see-ums (Ceratopogonidae) that seek blood meals at the expense of homeowners,
several factors make electric traps ineffective in reducing local mosquito popu-
lations (Surgeoner & Helson 1977, Nasci et al. 1983). Ultraviolet lamps that
emit considerable amounts of visible light (as do the lamps sold in commercial
electric traps) are less attractive to mosquitoes than lamps emitting only ultra-
violet wavelengths (Ikeuchi 1967). Furthermore, many species of mosquitoes
are not attracted to light traps at all (Pippin 1965, Miller et al 1969) and those
species that are are often not trapped in numbers proportionate to their popula-
tion sizes (Bradley 1943, Huffaker & Back 1943, Fox 1958). But perhaps the
most important reasons electric insect traps fail to reduce mosquito problems
are that 1) carbon dioxide exhaled by homeowners is far more attractive to
mosquitoes than are light traps (Headlee 1941, Huffaker & Back 1943, Nascit
et al. 1983), and 2) mosquitoes that do move toward traps are rarely killed by
electrocution devices (Service 1993).

1 Received August 18, 1995. Accepted September 2 1 , 1995.

2 Department of Entomology and Applied Ecology, Delaware Agricultural Experiment Station,
College of Agricultural Sciences, University of Delaware, Newark, Delaware 19717-1303.

ENT. NEWS 107(2): 77-82, March & April, 1996

78

ENTOMOLOGICAL NEWS

Electric insect traps are, however, effective at killing large numbers of non-
target insects. Nasci etal. (1983) found that the average zapper in South Bend,
Indiana killed more than 3000 insects per day, 96.7% of which were not female
mosquitoes. Little beyond ordinal totals is known about the diversity and sea-
sonal distribution of nontarget insects killed by zappers. As an initial step to-
ward understanding the ecological consequences of indiscriminant removal by
zappers of nontarget predators, parasitoids, and prey species from aquatic and
terrestrial ecosystems, we quantified at the family level the numbers and kinds
of insects killed over a season by homeowners’ zappers in a suburban setting.

MATERIALS AND METHODS

We asked six homeowners with active bug zappers in suburban Newark,
Delaware to participate in a summer-long study in 1994. All houses were within
3 km of a body of water. The house closest to water was about 65 meters from
a large stream containing many stagnant eddies. Another house abutted a wooded
area and was less than 1 km from a creek. The third house was about 1.5 km
from the same creek but farther upstream. The fourth was in a wooded cul-de-
sac through which ran a different creek; several permanent pools lay within 200
meters. The fifth house was situated in a residential development containing a
stream and scattered wooded areas; a small pond about 30 meters long and 15
meters wide was less than a kilometer away. A small stream about 3 km distant
was the nearest body of permanent water to the sixth house. Temporary pools,
tree holes and water-filled containers were scattered throughout the study area.
Thus, all traps were well within flight range of culicid and ceratopogonid breed-
ing sites.

From June 20 to July 9, 1994, homeowners were asked to run the traps one
night per week for at least two hours. Beginning on July 10, participants were
asked to run their zappers one night per week every other week for the nine
weeks ending August 27. A device constructed from a plastic dish 32 centime-
ters in diameter was suspended beneath each trap to collect electrocuted in-
sects. Each morning after the traps were run, we collected the samples from the
six sites and stored them in a freezer until they could be counted and identified
to family (except for Ephemeroptera,Psocoptera,Thysanoptera,and Trichoptera,
which were identified only to order, and several families of moths, which were
grouped as “Microlepidoptera”).

RESULTS

We collected 31 samples from the traps over our ten-week study period in
the summer of 1994. Nearly all electrocuted specimens, including the tiniest

Vol. 107, No. 2, March & April, 1996

79

Cecidomyiidae, were well-preserved and easily identified. Twelve orders and
more than 104 families were present in these samples and ranged in abundance
from a single individual (several families) to more than 4,600 individuals (Chi-
ronomidae; Table 1). Of the 13,789 insects killed by electric zappers in our
study, only 31 individuals (0.22%) were biting flies (female Culicidae, Simuli-
idae, and Ceratopogonidae). In contrast, insect predators, parasitoids, and
nonbiting aquatic insects were abundant (Table 1). Present in our counts were
representatives of 27 families of predators and nine families of parasitoids, to-
taling 1,868 individuals (13.5%). Carabid beetles, staphylinid beetles, cicadel-
lid leafhoppers, microlepidoptera, and braconid parasitoids were particularly
common victims. Large numbers of aquatic insects, such as caddisflies (Tri-
choptera) and midges (Chironomidae), were also destroyed; species from these
families represented nearly half (48.4%) of sample totals.

Average numbers of insects per trap declined sharply over the season (Fig.
1 ), ranging from a mean of 1 ,304 insects per trap on June 20 to just 1 06 insects
per trap on August 27. This probably reflects seasonal declines in the popula-

ei# c e c

June July August

Fig. 1. Seasonal pattern of insects killed at six electric insect traps in Newark, DEon six dales from
June 20 to August 27, 1994. Statistical interval = Standard Error. Pie charts depict the percentage of
the total catch consisting of nontarget insects (black portion) and biting flics (white portion) on
each trapping date

80

ENTOMOLOGICAL NEWS

tions of species attracted to these traps. Although biting insects generally in-
creased in proportion as the season progressed (from 0.26% of the total catch
on June 20 to 1 .88% on August 20), they still comprised a minuscule portion of
the total sample.

DISCUSSION

These data are straightforward: many thousands of nontarget insects repre-
senting a rich taxonomic diversity were destroyed by these traps. Only a tiny
fraction of trap victims were biting flies, the primary targets of electric zappers.
Since we did not independently measure mosquito populations in our study
sites we cannot definitively conclude that the zappers used in our study were
ineffective mosquito killers. However, three types of circumstantial evidence
suggest that this was indeed the case. First, it is highly unlikely that our low-
land, wooded sites which were rich in aquatic breeding habitats, produced so
few adult mosquitoes in the course of 9 weeks that 18 electrocuted females
would represent adequate control of these flies. Second, the preponderance of
aquatic insects in the samples suggests that our study traps were well within
the flight range of biting flies that breed in water (culicids, ceratopogonids).
Finally, our results are similar to those of Nasci et ai (1983) in which an inde-
pendent measure of culicid populations confirmed the inability of zappers to
attract mosquitoes that are present in suburban settings.

As we better understand the critical role insects play in the cohesion of
most non-marine ecosystems, the sale and use of electric insect traps that so
completely miss their advertised mark becomes increasingly irresponsible. It is
insects and other invertebrates, not vertebrates, that are the “glue” of ecosys-
tems; their elimination would inevitably lead to the rapid demise of those eco-
systems and their members, including Homo sapiens (Wilson 1987). Even if
targeted biting flies were effectively controlled by electric zappers, the result-
ing destruction of thousands of parasitoids, predators, aquatic insects, and other
members of the nocturnally active fauna would be difficult to justify.

Although we recognize its speculative shortcomings, a simple calculation
underscores the degree to which electric zappers may affect nontarget insect
populations. The seasonal mean catch per night (of at least 2 hr of trap time) as
quantified by our study totaled 445 insects per trap. Approximately one million
zappers are sold in the U.S. each year (Philadelphia Inquirer, 26 June 1995 p.
63). Electrocution devices are quite durable; the homeowners in our study had
been operating their units for an average of 7 yrs prior to 1994. If, in any given
year, 4 million traps are used for 40 nights during the summer, then
7 1 ,200,000,000 — more than 71 billion nontarget insects — are needlessly
destroyed in the U.S. each year by misinformed homeowners. If we substitute
into our calculations the trap means obtained by Nasci et ai (1983) in Indiana
(2163 insects during a 2 h trapping period; N = 10), this figure rises to nearly

Vol. 107, No. 2, March & April, 1996

Table 1. Seasonal totals of biting flies (in bold), predators and parasitoids (italicized), plus other
taxa killed by electric insect traps at six sites in Newark, DE.

No.

%of

No.

% of

Order and Family

Killed

Total

Order and Family

Killed

Total

Ephemeroptera . . .

15

0.11

Mycetophilidae . . .

...34

0.25

Dcrmaptera

Anisopodidae

... 13

0.09

Labiidae

2

0.02

Sciaridae

...89

0.65

Psocoptera

14

0.10

Dixidae

....3

0.02

Hemiptera

Cecidomyiidae . . . .

. . 316

2.29

Corixidae

10

0.07

Stratiomyidae

. . . . 5

0.04

Hebridae

2

0.02

Xylophagidae

.... 1

0.01

Miridae

89

0.64

Asilidae

. . . . 1

0.01

Nabidae

2

0.02

Scenopinidae

.... I

0.01

Lygaeidae

32

0.23

Rhagionidae

....2

0.02

Rhopalidae

I

0.01

Empididae

...58

0.42

Cydnidae

14

0.10

Dolichopodidae . . .

...70

0.51

Homoptera

Pipunculidae

....I

0.01

Cicadidae

33

0.24

Phoridae

...12

0.09

Cicadellidae

2421

17.56

Platypezidae

.... 4

0.03

Flatidae

8

0.05

Otitidae

.... 2

0.02

Acanaloniidae . . .

I

0.01

Tephritidae

2

0.02

Psyllidae

41

0.30

Sciomyzidae

.... 1

0.01

Delphacidae

1

0.01

Ephydridae

8

0.05

Cixiidae

1

0.01

Drosophilidae . . . .

....7

0.05

Aphididae

25

0.18

Agromyzidae

... 14

0.10

Thysanoptera ....

16

0.12

Lonchaeidae

5

0.04

Neuroptera

Lonchopteridae . . .

8

0.05

Corydalidae

I

0.01

Heleomyzidae . . . .

I

0.01

Chrysopidae

8

0.05

Sphaeroceridae . . .

2

0.02

Coleoptera

Anthomyiidae . . . .

... 28

0.20

Carabidae

661

4.79

Calliphoridae

... 17

0.12

Dytiscidae

21

0.15

Sarcophagidae . . . .

8

0.05

Hydrophilidae . . . .

83

0.60

Tachinidae

... 16

0.12

Staphylinidae

306

2.22

Trichoptera

, . 1597

11.58

Lucanidae

1

0.01

Lepidoptera

Scarabaeidae

219

1.58

Microlepidoptera . .

. 1121

8.13

Buprestidae

3

0.02

Tortricidae

, ... 19

0.14

Elateridae

46

0.33

Pyralidae

... 316

2.29

Lampyridae

12

0.09

Geometridae

... 35

0*25

Cantharidae

104

0.754

Lasiocampidae . . . .

3

0.02

Dermestidae

II

0.08

Arctiidae

... 11

0.08

Anobiidae

30

0.22

Noctuidae

...64

0.46

Cleridae

4

0.03

Notodontidae

2

0.02

Nitidulidae

27

0.20

Epipyropidae

....5

0.04

Coccinellidae . . . .

15

0.11

Yponomeutidae . . .

... . 10

0.07

Tenebrionidae . . . ,

13

0.09

Hvmenoptera

Mordellidae

10

0.07

Braconidae

. . 377

2.73

Cerambycidae . . .

11

0.08

Ichneumotiidae . . .

...77

0.56

Chrysomelidae . . ,

ii

0.16

Mymaridae

.... 1

0.01

Curculionidae . . . ,

7

0.05

Perilampidae

.... 1

0.0 1

Scolytidae

27

0.20

Eulophidae

.... 1

0.01

Diptera

Encyrtidae

.... 1

0.0 1

Tipulidae

223

1.62

Pteromalidae

.... 1

0.01

Psychodidae

11

Cf 25, 9 18

0.08

Tory mi da e

i

0.02

Culicidae

0.31

Eurytomidae

.... 1

0.01

Ccratopogonidae

Cf 30, 9 12

0.30

Chrysididae

....3

0.02

Chironornidae . . .

4612

33.45

Formicidae

... 84

0.61

Scatopsidae

13

0.09

Vespi

....3

0.02

Simuliidae

1

0.01

Halictidae

1

0.01

Bibionidae

1

0.01

82

ENTOMOLOGICAL NEWS

350 billion nontarget insects. We suggest, therefore, that while there is no evi-
dence that zappers control nuisance insects, their effects may be anything but
benign. Studies investigating the effects of insect defaunation on local ecosys-
tems in general and on specialized insectivores such as bats and nighthawks in
particular are needed to evaluate the ecological costs of zappers and other hu-
man activities destructive to insects. The results of our study indicate that ento-
mologists, especially those active in extension, should be educating the public
about the possible costs and lack of benefits from these gadgets.

ACKNOWLEDGMENTS

We gratefully acknowledge the Aftosmis, Fanny, Cherwaty, Walter, White, and Hawthorne
families, and C. J. Murphy and A. Schleiniger for their cooperative participation in our study. R. R.
Roth, H. Frick, R. G. Weber, and C. Tallamy made helpful comments on the manuscript. Published
as Paper No. 1557 of the Delaware Agricultural Experiment Station; Contribution No. 678 of the
Department of Entomology and Applied Ecology.

LITERATURE CITED

Bradley, G. H. 1943. Determination of densities of Anopheles quadrimaculatus on the wing. Proc.
New Jersey Mosq. Exterm. Assoc. 30:22-27.

Fox, I. 1958. The mosquitoes of the international airport, lsla Verde, Puerto Rico, as shown by light
traps. Mosq. News 18:1 17-124.

Hcadlee, T. J. 1941. New Jersey mosquito problems. Proc. New Jersey Mosq. Exterm. Assoc.
28:7-12.

Huffaker, C. B., and R. C. Baek. 1943. A study of methods of sampling mosquito populations. J.
Econ. Entomol. 36:561-569.

Ikeuehi, M. 1967. Ecological studies on mosquitoes collected by light traps. Trop. Med. 9: 186-

200 .

Miller, T. A., R. G. Stryker, R. N. Wilkinson, and S. Esah. 1 969. Notes on the use of CCb-baited
CDC miniature light traps for mosquito surveillance in Thailand. Mosq. News 29:688-689.
Nasci, R. S., C. W. Harris, and C. K. Porter. 1983. Failure of an insect electrocuting device to
reduce mosquito biting. Mosq. News 43:1 80-44.

Pippin, W. F. 1965. Notes on the operation of a light trap in central Luzon, Philippine Islands.
Mosq . News 25: 1 83- 1 87.

Service, M. W. 1993. Mosquito ecology: field sampling methods. 2nd edition. Elsevier Applied
Science, New York, New York.

Surgconer, G. A., and B. V. Helson. 1 977. A field evaluation of electrocutors for mosquito control
in southern Ontario. Proc. Entomol. Soc. Ontario 108:53-58.

Wilson, E. O. 1987. The little things that run the world (the importance and conservation of
invertebrates). Conservation Biology 1:344-346.